Vitamin A

TABLE 3.2Chemical Nomenclature and Provitamin A Activity of some Common Carotenoids of Plant Foods Trivial Name Semisystematic Name Type Vitamin A Activity % a Lycopene c,c-Carotene Acyc

Retinol derived from ingested provitamin A carotenoids, along with thatingested as such, is stored in the liver and secreted into the bloodstreamwhen needed The circulating retinol is taken up by target cells andconverted in part to retinoic acid, which functions as a ligand to anuclear retinoid receptor The liganded receptor interacts with specificenhancer sites on the DNA and, in collaboration with many other regulat-ory proteins, induces the synthesis of proteins through the direct control

of gene transcription This type of action establishes vitamin A (in theform of the retinoic acid metabolite) as a hormone, similar to the steroidhormones and thyroid hormone Vision is a nonhormonal, biochemicalprocess involving a different vitamin A metabolite, 11-cis-retinaldehyde.Vitamin A is an essential dietary factor for normal embryogenesis, cellgrowth and differentiation, reproduction, maintenance of the immunesystem, and vision Malnourished children in famine-stricken countriesare at risk of clinical vitamin A deficiency, which manifests as keratiniza-tion of the conjunctiva, and later of the cornea, causing permanentblindness There is also increased infant mortality from infectious dis-eases due to impaired immunocompetence

Excessive dietary intakes of vitamin A produce symptoms of acuteand chronic toxicity, including teratogenicity in developing fetuses.Normally, toxicity results from the indiscriminate use of pharmaceuticalsupplements, and not from the consumption of usual diets The onlynaturally occurring products that contain sufficient vitamin A to inducetoxicity are the livers of animals at the top of long food chains, such aslarge marine fish and carnivores

Carotenoids are widely believed to protect human health In particular,some epidemiological studies have correlated the intake of carotenoid-rich fruits and vegetables with protection from some forms of cancer,cardiovascular disease, and age-related macular degeneration This action

is not restricted to provitamins and, therefore, may be attributable to thecarotenoids’ antioxidant properties rather than to their vitamin A activity

3.2 Chemical Structure, Biopotency, and Physicochemical Properties

3.2.1 Structure and Biopotency

3.2.1.1 Retinol

The structures of retinoids found in foods and fish-liver oils are shown inFigure 3.1 The parent vitamin A compound, retinol, has the empirical

comprises a b-ionone (cyclohexenyl) ring attached at the carbon-6 (C-6)position to a polyene side chain whose four double bonds give rise to cis–trans (geometric) isomerism Theory predicts the existence of 16 possibleisomers of retinol, but most of these exhibit steric hindrance, and someare too labile to exist [1] The predominant isomer, all-trans-retinol,

13 14

15 16

18

9 9

10

10

13 14

13 14

possesses maximal (100%) vitamin A activity and is frequently panied in natural foodstuffs by smaller amounts of the less potent 13-cis-retinol Lower potency 9-cis-retinol and 9,13-di-cis-retinol occur in small

major form of vitamin A in the liver and flesh of freshwater fish

A content of foodstuffs

The biopotencies of isomers of vitamin A esters determined by means

of rat vaginal smear assays are presented in Table 3.1 [2] Retinaldehydepossesses about 90% of the biological activity of all-trans-retinol and3-dehydroretinol is about 40% as active [3]

3.2.1.2 Provitamin A Carotenoids

Carotenoids can be considered chemically as derivatives of lycopene — a

C40H56

abbreviation ip for the isoprenoid unit, the carotenoids can be represented

as ip-ip-ip-ip-pi-pi-pi-pi, that is, the arrangement of the units is reversed

at the center of the molecule Derivatives are formed by a variety of tions that include cyclization, hydrogenation, dehydrogenation, andinsertion of oxygen Hydrocarbon carotenoids are known as carotenes,and the oxygenated derivatives are termed xanthophylls The oxygenfunctions of xanthophylls are most commonly hydroxy, keto, epoxy,methoxy, and carboxy groups Some acyclic carotenoids occur widely,for example, lycopene, but monocyclic and bicyclic compounds aremore common Most carotenoids of plant tissues contain 40 carbonatoms, but shortened molecules known as apocarotenoids can arise as aresult of partial oxidative cleavage

A semisystematic nomenclature for carotenoids has been devised toconvey structural information [4] According to this scheme, the caroten-oid molecule is considered as two halves, and the nature of the end group

of each half is specified Each carotenoid is considered to be a derivative

of a parent carotene, indicated by two Greek letters describing theend groups The nomenclature recognizes six end groups: b (beta),

hydrogenation level and the presence of substituent groups are indicated

by the use of conventional prefixes and suffixes The numbering systemfor carotenoids is shown in the structure of b-carotene (Figure 3.2).Many of the naturally occurring carotenoids are chiral, bearing one tofive asymmetric carbon atoms In most cases, a given carotenoid occurs

in only one chiral form The absolute configuration at a chiral center isdesignated by use of the R,S convention Unless stated otherwise,all double bonds have the trans configuration Cis –trans isomerism isindicated by citing the double bond or bonds with a cis configuration

13 14 15

15 15´

The Z, E terminology for geometric isomerism is seldom used in vitamin

A and carotenoid nomenclature The semisystematic names of someFrom a nutritional viewpoint, the carotenoids are classified as provita-mins and inactive carotenoids To have vitamin A activity, the carotenoidmolecule must incorporate a molecule of retinol, that is an unsubstituted

of two molecules of retinol joined tail to tail; thus the compound possessesmaximal (100%) vitamin A activity The structures of all other provitamin

A carotenoids incorporate one molecule of retinol and hence theoreticallycontribute 50% of the biological activity of b-carotene Among the 600 or

so carotenoids that exist in nature, only about 50 possess vitamin Aactivity in varying degrees of potency

In nature, carotenoids exist primarily in the all-trans configuration, butsmall amounts of 9-cis, 13-cis, and 15-cis isomers of b-carotene have beenfound in fresh and processed fruits and vegetables [6,7,7a,8] With asym-metrical carotenoids, such as a-carotene and b-cryptoxanthin, the number

of theoretically possible cis isomers is approximately twofold greater thanwith symmetrical carotenoids, such as b-carotene In fruits, hydroxycaro-tenoids (carotenols) exist mainly as mono or bis esters of saturated long-chain fatty acids, such as lauric (C12), myristic (C14), and palmitic (C16)acids [9,10]

nonpro-vitamin carotenoid canthaxanthin are permitted food color additives [11].13-cis-b-Carotene and 9-cis-b-carotene exhibit, respectively, 53 and 38%

of the provitamin A activity of all-trans-b-carotene in a rat growthassay [12]

3.2.2 Physicochemical Properties

3.2.2.1 Appearance and Solubility

Retinol and retinyl acetate are yellow crystalline powders; retinyl tate is a pale yellow oil or crystalline mass b-Carotene is a reddish-brown

rust-red crystalline powder

Retinol and its esters are insoluble in water; soluble in alcohol; andreadily soluble in diethyl ether, petroleum ether, chloroform, acetone,and fats and oils b-Carotene is insoluble in water; very sparinglysoluble in alcohol, fats and oils; sparingly soluble in ether and acetone;and slightly soluble in chloroform

TABLE 3.2

Chemical Nomenclature and Provitamin A Activity of some Common Carotenoids of Plant Foods

Trivial Name Semisystematic Name Type

Vitamin A Activity (%) a

Lycopene c,c-Carotene Acyclic carotene Inactive

z-Carotene 7,8,70,80-Tetrahydro-c,c-carotene Acyclic carotene Inactive

g-Carotene b, c -Carotene Monocyclic carotene 42–50

b-Zeacarotene 70,80-Dihydro-b, c -carotene Monocyclic carotene 20–40

b-Carotene b,b-Carotene Bicyclic carotene 100

a-Carotene (60R)-b, 1 -Carotene Bicyclic carotene 50–54

b-Cryptoxanthin (3R)-b,b-Caroten-3-ol Bicyclic, monohydroxy-carotenoid 50–60

Zeinoxanthin (3R,60R)-b, 1 -Caroten-3-ol Bicyclic, monohydroxy-carotenoid Inactive

Zeaxanthin (3R,30R)-b,b-Carotene-3,30-diol Bicyclic, dihydroxycarotenoid Inactive

Lutein (3R,30R, 60R)-b,1-Carotene-3,30-diol Bicyclic, dihydroxycarotenoid Inactive

b-Carotene-5,6-epoxide 5,6-Epoxy-5,6-dihydro-b,b-carotene Bicyclic, monoepoxy-carotenoid 21

7-didehydro-5,6,50,60-tetrahydro-b, b-carotene-3,5,30-triol

Bicyclic, monoepoxy-, trihydroxycarotenoid

Inactive

Violaxanthin (3S,5R, 6S,30S, 50R,60S)-5,6,50,

60-Diepoxy-5,6,50,60-tetrahydro-b, b-carotene-3,30-diol

Bicyclic, diepoxy-, dihydroxycarotenoid Inactive

a

Activity of all-trans forms relative to the activity of b-carotene.

Source: From Bauernfeind, J.C., J Agric Food Chem., 20, 456, 1972 With permission.

3.2.2.2 Stability in Nonaqueous Solution

Retinol is readily oxidized by atmospheric oxygen, resulting in an almostcomplete loss of biological activity The 5,6-epoxide and 5,8-furanoxideare among the oxidation products Retinyl esters are somewhat morestable towards oxidation than retinol Retinol is extremely sensitivetowards acids, which can cause rearrangement of the double bonds anddehydration Solutions of all-trans-retinol or retinyl palmitate in hexaneundergo slow isomerization to the lower potency cis isomers whenexposed to white light The photoisomerization rate is greatly increased

in the presence of chlorinated solvents, but under gold fluorescent light(wavelengths greater than 500 nm) no significant isomerization occurswithin 23 h [13] Retinyl palmitate is stable in chlorinated solventswhen it is stored in the dark [14] Irradiation also rearranges doublebonds to form inactive retro structures, which are responsible for much

of the yellow color of decomposing vitamin A [15]

The carotenoids are stable within their natural plant cell environment,but once isolated they are prone to molecular rearrangement, trans tocis isomerization, and degradation by heat, light, oxygen, trace amounts ofacids, and active surfaces such as silica Xanthophylls are particularly sus-ceptible to these agents and are also destroyed in alkaline environments.Chlorophyll compounds naturally present in extracts of green leafyvegetables have the ability to sensitize the photoisomerization of caroten-oids, giving rise to appreciable amounts of cis isomers during even abrief exposure to white light [16] Solutions of b-carotene undergo slowisomerization, giving rise to 9-cis and 13-cis isomers, even when stored inthe dark In general, isomerization is higher in nonpolar solvents than inpolar solvents [17] Chlorinated solvents are often contaminated with traceamounts of hydrochloric acid, which can promote stereoisomerization

3.3 Vitamin A in Foods

3.3.1 Occurrence

All natural sources of vitamin A are derived ultimately from provitamin

A carotenoids, which are synthesized for metabolic purposes exclusively

by higher plants and photosynthetic microorganisms Meat and milkcontain vitamin A as a consequence of the animal converting ingestedprovitamin A carotenoids to retinol Carotenoids present in milk

products, egg yolk, shellfish, and crustacea result from the deposition ofunmetabolized dietary carotenoids in the animal’s tissues In green,photosynthetic plant tissue, carotenoids are localized with the chloro-phylls in the thylakoid membranes of chloroplasts, bound noncovalently

to proteins in pigment—protein complexes In nonphotosynthetic tissues(e.g., in fruits, carrots, and sweet potatoes), carotenoids are primarilyfound in chromoplasts, either within lipid droplets or associated withproteins, depending on the chromoplast type [18] Carotenoids alsooccur as very fine dispersions in aqueous systems, such as orange juice.Foods are supplemented with vitamin A in the form of standardizedpreparations of synthetic esters of retinol, nowadays chiefly retinyl palmi-tate The preparations are available commercially as either dilutions inhigh-quality vegetable oils containing added vitamin E as an antioxidant

or as dry, stabilized beadlets, in which the vitamin A is dispersed in asolid matrix of gelatin and sucrose or gum acacia and sucrose The oilypreparations are used to supplement fat-based foods such as margarines.The dry preparations are used in dried food products such as milkpowder, infant formulas, and dietetic foods b-Carotene, in the form ofmicrocrystals suspended in vegetable oil, is used to impart color to fat-based foods such as margarines, butter, and cheese Dried emulsions ofcarotenoids can be rehydrated and used to color and nutrify a variety

of water-based foods [19]

3.3.1.1 Vitamin A

[20] The liver of meat animals is a rich source of vitamin A, for this organ

is the body’s main storage site of the vitamin Whole milk, butter, cheese,and eggs are important dietary sources Margarine is fortified withvitamin A to make it nutritionally equivalent to butter Fortification ofskim milk, partially skimmed milk, and nonfat dry milk with vitamin A

is mandatory in the U.S and Canada The edible portions of fatty fish(e.g., herring, mackerel, pilchards, sardines, and tuna) contain moderateamounts of vitamin A, but white fish, apart from the haddock, containonly trace amounts In most of the foods that contain vitamin A, theretinol forms esters with long-chain fatty acids, particularly palmiticacid An exception is egg yolk, in which unesterified retinol representsthe major retinoid, accompanied by retinaldehyde and retinyl esters[21] Cis isomers of vitamin A occur in foods to varying extents, withfish-liver oils and eggs containing as much as 35 and 20%, respectively,

of their total retinol in this form [3]

In margarines, the various naturally occurring vitamin A isomers andprovitamins that may have been present in the original crude oils areremoved during the refining process [22] Thus the only vitamin A that

is present in the final margarine is the retinyl ester and b-carotenecombination that is added during production [23].

3.3.1.2 Provitamin A Carotenoids

Carrots, sweet potatoes, and green leafy vegetables are major contributors

of provitamin A in the American diet [24] Carotenoid concentrations offruits and vegetables are affected by factors such as: (1) cultivar/variety;(2) part of the plant consumed; (3) uneven distribution of the carotenoids

in a given food sample; (4) stage of maturity; (5) climate/geographic site

of production; (6) harvesting and postharvest handling; and (7) sing and storage [25] Compared to vegetables, fruits contain a greatervariety of carotenoids in varying concentrations Citrus fruits are themost complex fruits in terms of the number of carotenoids found Inripening fruits, the decrease in chlorophylls is frequently accompanied

proces-by an increase in the concentration of carotenoids and an increase in the

TABLE 3.3

Vitamin A Content of Various Foods

Food

Micrograms of Retinol per 100 g Edible Portion

Note: Tr, trace; N, the vitamin is present in significant quantities but there is no reliable information on the amount.

Source: From Food Standards Agency, McCance and Widdowson’s The sition of Foods, 6th summary ed., Royal Society of Chemistry, Cambridge,

Compo-2002 With permission.

ratio of carotenes to xanthophylls Red palm oil is an important foodsource of vitamin A in South America, Southeast Asia, and some countries

of Africa This oil is extracted from the fleshy mesocarp of the palm nut;the oil extracted from the palm kernel is without value as a source ofvitamin A [26] Cereals (apart from yellow maize) are negligible sources

caroten-of the vitamin A compounds

The many studies of the effects of processing and domestic cooking onthe levels of carotenoids in foods have produced conflicting results, owing

to differences in the experimental approach [33] One of the factorsthat may lead to considerable variation in analytical data on raw and

TABLE 3.4

Quantitative Distribution of b-Carotene, a-Carotene, and b-Cryptoxanthin Isomers in Fresh and Processed Fruits and Vegetablesa,b

Extract

All-trans 9-cis 13-cis 15-cis other cis c Total All-trans 9-cis 13-cis 130-cis other cis d Total All-trans 13/130-cis 15-cis Total Total e Broccoli

TABLE 3.4 Continued

Extract

All-trans 9-cis 13-cis 15-cis other cis c Total All-trans 9-cis 13-cis 130-cis other cis d Total All-trans 13/130-cis 15-cis Total Total e Spinach

a Data based on an average of two lots, except for orange juice, which had only one lot Each lot was analyzed by HPLC in duplicate.

b Concentrations are in micrograms per gram of dry weight tissue.

c Totals the concentration of one unidentified cis isomer of b-carotene.

d Totals the concentration of two unidentified cis isomers of a-carotene.

e Total provitamin A carotenoid concentration includes all isomers of b-carotene, a-carotene, and b-cryptoxanthin.

cooked vegetables is the incomplete extraction of protein-bound oids from the raw samples Other factors are unaccounted loss of waterand leaching of soluble solids The most reliable data are provided bywell-designed protocols that employ high-performance liquid chromato-graphy (HPLC) to determine individual carotenoids.

caroten-3.3.2.2 Vitamin A in Milk

Low-fat fluid milks are fortified commercially by adding small quantities

of a carrier oil containing an emulsifier and retinyl palmitate [34] Morethan 99% of the native vitamin A in milk occurs as retinyl esters of satu-rated and unsaturated fatty acids; the remaining fraction occurs as freeretinol The distribution of the predominant retinyl esters in pasteurizedbovine milk, expressed as percent retinol equivalents, is as follows: palmi-tate (16:0), 36.7%; oleate (18:1), 20.3%; linolenate (18:3), 8.5%; stearate(18:0), 8.4%; linoleate (18:2), 7.3%; pentadecanoate (15:0), 6.0% [35].Vitamin A stability in milk depends on the method of processing, the fatcontent, whether or not the product is fortified with vitamin A, the packa-ging, and the holding conditions prior to consumption Dissolved oxygenand oxygen in the container’s headspace adversely affect the stability ofvitamin A [36] During heat processing and storage of milk, the vitamin

A is potentially subject to trans – cis isomerization as well as oxidation.Cis isomers of vitamin A exhibit reduced biopotency relative to theall-trans form and, therefore, the analytical method employed for deter-mining vitamin A must be capable of distinguishing between cis andall-trans forms This can be achieved by means of normal-phase HPLC.Saponification simplifies the analysis by hydrolyzing the various esters

to retinol The loss of ester information is of no practical consequence as

it has been demonstrated that all the retinyl ester species have similarstability in milk fat [35]

Panfili et al [37] determined the degree of isomerization of vitamin A invarious dairy products marketed in Italy In unprocessed “raw” milk,there was no conversion of all-trans retinol to cis forms Pasteurizedmilk treated for 15 s at temperatures ranging from 72 to 768C had anaverage 13-cis:all-trans ratio of 6.4% Ratios for ultra-high temperature(UHT) and sterilized milks were 15.7 and 33.5%, respectively, consistentwith increased isomerization at the higher processing temperatures Theheating of pasteurized whole milk in a household microwave ovendoes not cause loss of vitamin A [38]

UHT milk and powdered milk can be stored in a safe edible state forlong periods at ambient temperature, provided that the packaging isimpervious to moisture, air, and light It is important to recognize that

“ambient temperature” can be higher than 20– 258C in some parts of theworld and that the rate of biochemical reactions depends largely on the

storage temperature During storage of UHT milk, the main cause ofeventual spoilage is proteolysis and gelation [39] Milk powder is fre-quently consumed in developing countries after long storage periods attropical temperatures The keeping quality of milk powder is dependent

enzymic activity [40]

Woollard and Edmiston [41] monitored the stability of vitamin A in tified whole and skim milk powder stored at ambient temperature inheat-sealed, laminated aluminum foil sachets Six-monthly checks weremade over a 2-year period Losses of vitamin A in six whole milkpowder samples after 6 months ranged from 24 to 50%; loss in anothersample was 2% After 24 months, losses in the whole milk powdersamples were not much greater (24 – 66%) In contrast, the two samples

for-of skim milk powder tested lost 93% and 95% vitamin A after 6 months,and 100% after 18 months

Woollard and Fairweather [42] studied the storage stability of vitamin

A in 2% low-fat UHT milk sampled at the time of manufacture TheUHT treatment was an indirect heating process (1388C for 4 s) and themilk was packaged aseptically in 250-ml Tetrapak cartons formed frompolyethylene-coated paper The use of normal-phase HPLC on unsaponi-fied samples allowed natural (palmitate) and supplemental (acetate)vitamin A esters to be determined independently Losses of bothnatural and added vitamin A occurred with time In two trials atambient temperature, total vitamin A decayed progressively up to 15weeks of storage, when the losses were 33 and 45% The respectivelosses at the product’s 28-week expiry date were 35 and 47% At simu-lated tropical temperature (358C), the total losses of vitamin A in twotrials increased to 53 and 61% at 28 weeks

Woollard and Indyk [43] studied the effects of processing and storage

on vitamin A isomerization in UHT milk and milk powders They firststudied retinyl ester isomerization in vitamin A-fortified 2% low-fatUHT milk stored at 208C for up to 8 months The UHT treatment wasindirect and the milk was packaged in 250-ml Tetrapak cartons Cisisomers were undetectable in raw milk, but appeared after UHT treat-ment Both supplementary and native all-trans esters showed degradationwith time, but the cis isomers maintained their concentrations until theeventual onset of milk spoilage The next experiment was to comparethe effects of direct and indirect UHT treatment on two identical skimmilk samples fortified with retinyl acetate After 3 days of storage, survi-val of all-trans retinyl acetate in the directly heated sample was markedlyhigher than in the indirectly heated sample However, the cis isomer con-centrations (mainly 13-cis retinyl acetate) differed little between the twomilks Fortified whole milk powders sealed in aluminum foil sachets

analyzed as soon as possible after production, except for one sample (ratio0.42) where an abnormally poor quality vitamin premix was used Thecis:trans ratio gradually increased during storage of the milk powder for

up to 3 years Storing canned fortified whole milk powders at 378C and458C for 3 months caused significant temperature-related increases inall-trans vitamin A degradation and concomitant increases in cis isomerconcentration At 278C, there were no significant changes from initialconcentrations

Vidal-Valverde et al [44] studied the effects of storage temperature andtime on the retinol content of four commercial unfortified whole UHTmilk samples, two directly heated and two indirectly heated Analysisinvolved saponification of samples and reversed-phase HPLC Significantlosses of retinol were observed after 1 month of storage at 308C, generallyincreasing with storage time Increasing the storage temperature from

30 to 408C had a variable effect, depending on the particular milk.Frozen storage (2208C) for up to 60 days had no effect on the retinol

of 17 –18%

Albala´-Hurtado et al [45] assessed the extent of vitamin A loss in threeliquid milk-based infant formulas stored for 1 year at 20, 30, and 378C.Two samples (brands A and B) were “follow-on” milks for infants agedfrom 4– 6 months to 12 months, and the other a “junior milk” (brand B)for infants after the first year of life The expiry date of these productswas 6 months Vitamin A remained stable in the brand A follow-onmilks stored for 12 months at 20 and 308C, but there were significant

end of this period In the brand B follow-on and junior milks there were

the greatest losses being at 378C (57–69%) In a comparison of liquid andpowdered milk-based infant formulas claimed to have the same finalcomposition, contents of vitamin A throughout storage at 20 and 308Conly showed a slight decrease, which was not statistically significant

At 378C, losses after 12 months of storage were practically identical inpowdered (36%) and in liquid (34%) samples However, the decreaseoccurred earlier in powdered samples than in liquid samples [46].The presence of fat in milk protects vitamin A from degradation, asdemonstrated by Woollard and Edmiston [41] in stored fortified milkpowders Lau et al [47] prepared milk samples with 0.15, 3, 6, and 10%fat and found that the rate of degradation of added retinyl palmitate inbeadlet form increased with decreasing fat content Final vitamin A con-centrations at the end of 3 weeks of storage were higher in milk with highfat and closely corresponded to the native vitamin A concentrationspresent in the 3-, 6-, and 10%-fat milks Natural vitamin A, being localized

in the fat globules, would be protected from oxygen On the other hand,

the added retinyl palmitate, being dispersed in the serum phase, could bepreferentially oxidized due to greater contact with oxygen Cox et al [48]reported that the added vitamin A (retinyl palmitate dissolved in glycerylmono-oleate) is more stable in dry whole milk than in nonfat dry milk.Whereas the added vitamin A is likely to be dissolved in the fat phase

of dry whole milk, the vitamin added to nonfat dry milk would besubject to more surface exposure on nonfat material

Some studies of UHT milk have shown an association between vitamin

A degradation and residual oxygen in the sample [36,49], pointing to dation as one of the mechanisms involved In addition to dissolvedoxygen, the antioxidant NDGA effectively retarded vitamin A degra-dation in fluid milks and air-packed dry milks [48] UHT milk that hasbeen processed by direct heating has a very low oxygen content,whereas the indirect process, when employed without a deaerator,gives a product that may be near saturated in oxygen [50] Vitamin Ahas been shown to be more stable in directly heated than in indirectlyheated UHT milk [43], which is consistent with oxidation being partlyresponsible for loss of the vitamin Le Maguer and Jackson [36] observedthat rapid degradation of vitamin A in UHT milk only occurred after adelayed period (12 – 16 weeks) and suggested that the initial lag phasemay be due to the effective scavenging of residual oxygen by free sulfhy-dryl groups made available as a result of the heat treatment upon wheyproteins, chiefly b-lactoglobulin [51] In support of this suggestion,vitamin A destruction was less in dried milks which had received apreheat treatment at 82.28C for 30 min when compared with a preheattreatment at 62.88C for 30 min [48]

oxi-A distinct hay-like off-flavor is frequently detectable in vitamin oxi-fortified pasteurized low-fat milk that has been stored in a refrigeratorfor several days The hay-like flavor is definitely associated with thepresence of added vitamin A in the milk, because it does not occur in non-fortified milk [52] In a study of fortified nonfat dry milk [53], the intensity

A-of the hay-like flavor became weaker as the level A-of milk fat increased,thus the milk fat masks the off-flavor The incidence of hay-like flavorincreased as the vitamin A levels decreased during storage, suggestingthat the off-flavor was a chemical product(s) of vitamin A decomposition.The addition of ascorbic acid prevented the degradation of vitamin A inskim milk stored at 48C, indicating that oxidation of vitamin A was thelikely cause of hay-like flavor development Two major oxidation pro-ducts of retinyl palmitate, namely b-ionone and dihydroactinidiolide,have a mild hay-like odor in the pure state and may be responsible forthe flavor defect [54] The best fortification methods for preventingvitamin A degradation and hay-like flavor production in nonfat drymilk were dry-blending with vitamin A beadlets or addition of beadlets

at the agglomeration chamber during instantization [53]

Milk in a transparent container is inevitably exposed to light duringprocessing, handling and storage Exposure of milk to light causes apartial destruction of vitamin A, which increases as the fat contentdecreases (Table 3.5) [55] The addition of 3% nonfat milk solids to skimmilk protects the retinyl palmitate added to the milk [56] Retinyl palmi-tate added to homogenized whole milk is more susceptible to photodes-truction than indigenous retinyl esters [57] Aqueous-based vitamin Aconcentrate provided more stability to fluorescent light in 2% low-fatmilk than oil-based concentrate, while the reverse occurred in skimmilk [58] The type of carrier oil influences the stability of addedvitamin A toward light; butter oil and coconut oil were much more pro-tective than corn oil or peanut oil [59] No correlation was foundbetween peroxide value of the carrier oil and vitamin A loss [60], aswould be expected if degradation was a consequence of lipid peroxi-dation Photodegradation of added vitamin A was not prevented by theaddition of NDGA, although this antioxidant effectively retardedvitamin A degradation in all air-packed milk samples tested [48] Theselatter two observations imply that the destructive effect of light onvitamin A in milk is not due to autoxidation, at least not entirely Whenskim milk fortified with retinyl palmitate is exposed to fluorescent light,the concentration of all-trans retinyl palmitate decreases along with theconcentration of the minor 13-cis isomer; this is accompanied by theappearance of 9-cis retinyl palmitate [61,62] Jung et al [62] reportedthat addition of ascorbic acid to the fortified skim milk inhibited theloss of all-trans and 13-cis retinyl palmitate, but increased the formation

of the 9-cis isomer It appeared that exposing the milk to light inducedisomerization of all-trans retinyl palmitate to the 9-cis form as well asoxidation of the isomers Ascorbic acid decreased oxidation of retinylpalmitate isomers, but did not prevent isomerization of all-trans to 9-cis.The photodegradation of riboflavin in milk may be important in the

TABLE 3.5

Effect of Fat Content on Loss of Added Retinyl Palmitate in MilkExposed to Fluorescent Light

Percentage Loss of Vitamin A

loss of vitamin A because the decrease in all-trans retinyl palmitate wasmore pronounced when 3 mg/ml riboflavin was added to the skimmilk Addition of ascorbic acid reduced this loss Photodegradation ofriboflavin in milk generates superoxide anion radical and singlet oxygen,two highly reactive oxygen species [63] The results suggest that thesereactive species are, to some extent, involved in the light-induced destruc-tion of all-trans retinyl palmitate Ascorbic acid is an effective scavenger

of all aggressive reactive oxygen species [64] and conjecturally, by ing this role when added to milk, it reduces the light-induced loss ofvitamin A to a considerable extent Fortified skimmed milk in glass orplastic containers showed similar losses of all-trans and 13-cis retinylpalmitate when exposed to 3 days of continuous illumination No loss ofeither isomer occurred in paperboard cartons [65]

perform-3.3.2.3 Supplemental Vitamin A in Corn Flakes and Rice

Kim et al [66] studied the degradation of retinyl palmitate added to corn

D) during storage of the flakes at ambient temperature (average 238C) and

at elevated temperature (458C) Individual 400-g samples were stored inthe original sealed plastic liner and cardboard box packaging system,protected from light The initial isomer composition of the retinyl

dis-tribution of these isomers remained nearly constant throughout storage,irrespective of sample type and storage conditions After 6– 8 weeks ofstorage, about 90% of the retinyl palmitate was lost in all samples,except those fortified with the vitamin mixture and kept at ambient temp-erature In the latter samples, about 30 –40% of retinyl palmitate was lostafter the same period Thus the presence of other vitamins reduced theloss of retinyl palmitate from corn flakes stored at ambient temperature,but not at elevated temperature

Rice is an ideal candidate for vitamin A fortification because vitamin Adeficiency is prevalent in certain countries where rice is a staple food A

Washington) has been developed that overcomes the problem of loss ofthe fortificant during the typical washing and cooking methods UltraRice is made from broken rice grains, which typically comprise 20– 30%

of the harvest The grains are milled into rice flour, combined with abinder and retinyl palmitate, and reformed into rice grains with similarappearance and texture to whole rice grains This preparation is thenused as a premix for blending with virgin, unfortified rice at a ratio of1:100 to 1:200 to yield a final product that can provide 100% of the dailyvitamin A requirement for a child eating a typical daily quantity of rice.Lee et al [67] mixed Ultra Rice with a long grain rice at the ratio of

1:100 and studied the stability of retinyl palmitate in the mixture duringcooking and storage for 6 months After cooking, 75– 87% of the retinylpalmitate was retained, depending upon the cooking methods A signifi-

stored at 23 and 358C was observed, but no difference in the stabilitywas found between 55 and 80% relative humidity at the two tempera-tures When stored at 238C for 6 months, 85% of the retinyl palmitatewas retained compared with about 50% retention at 358C The producthalf-life (time required to reach 50% loss of retinyl palmitate) at differentstorage conditions is given in Table 3.6 The results show that storageunder tropical conditions results in extensive losses of vitamin A

3.3.2.4 Provitamin A Carotenoids

Being highly unsaturated, carotenoids are prone to trans –cis tion and oxidation, resulting in loss of color and provitamin A activity.The major cause of carotenoid destruction during processing andstorage of foods is enzymatic or nonenzymatic oxidation The enzymelipoxygenase in plant tissues catalyzes lipid peroxidation, giving rise tohydroperoxides The hydroperoxides decompose to form peroxyl andalkoxyl radicals which attack carotenoids The cutting of fruits and veg-etables into small pieces or maceration increases exposure to oxygenand brings the carotenoids and enzymes together Exclusion of oxygen,

isomeriza-by vacuum- or hot-filling, impermeable packaging, or inert atmosphere,diminishes carotenoid decomposition during storage

Moderate heat treatments such as blanching and cooking denaturecarotenoid binding proteins, thereby releasing the carotenoids so thatthey can be more readily extracted Blanching of fruits and vegetablesbefore processing inactivates lipoxygenase and other enzymes (e.g., per-oxidase) that are involved in carotenoid destruction Steam blanching

TABLE 3.6

Half-Life of Retinyl Palmitate in Rice Fortified with Ultra

RicewUnder Different Storage Conditions

Note: RH, relative humidity.

Source: From Lee, J., Hamer, M.L., and Eitenmiller, R.R., J Food

Sci., 65, 915, 2000 With permission.

increased the b-carotene concentration of broccoli and carrots relative toraw concentrations and only slightly decreased the b-carotene concen-tration of green beans [68] Overcooking can cause loss of provitamin Aactivity due to the action of heat and oxygen to form cis isomers andepoxy-carotenoids Optimal retention of carotene is obtained by steamingvegetables or cooking with minimal water until the vegetables are cookedbut still crisp [69] Khachik et al [70] examined the effect of variouscooking methods on the carotenoid concentration of green vegetablesand tomatoes determined by HPLC Samples were taken from severalsources and the carotenoid content in the cooked vegetables and tomatoeswas compared to that in raw samples of equal weight taken from the samestarting material Care was taken to ensure the complete extraction of theprotein-bound carotenoids from the raw samples The cooking conditionsand concentrations of a- and b-carotene in broccoli, spinach, green beans,and tomatoes are shown in Table 3.7 The moderate heat treatments ofsteaming, microwaving, 9-min boiling, and stewing did not significantlyaffect the concentrations of a- and b-carotene; surprisingly, neither did themore severe treatment of boiling green beans for 1 h.

TABLE 3.7

Effect of Cooking on Carotene Content of Green Vegetables and Tomatoes

mg Carotene per/100 g of Edible Food (Raw) (mean + S.D.)

Note: nd, not detected.

a Mean values are for three batches.

b

Mean values are for four batches.

c

Mean values are for two batches.

Source: From Khachik, F., Goli, M.B., Beecher, G.R., Holden, J., Lusby, W.R., Tenorio, M.D., and Barrera, M.R., J Agric Food Chem., 40, 390, 1992 With permission.

Direct sun exposure of leafy vegetables, as often practiced in the tropics,results in considerable destruction of total carotene, but retention isimproved by minimizing light exposure through shade provision [71].Dehydrated plant tissues undergo oxidation of carotenoids due to theincreased surface area exposed to oxygen The loss of water furtherreduces carotenoid stability through the relationship between wateractivity and lipid oxidation (Section 1.5.1) Carotenoid content in dehy-drated carrots, broccoli, and spinach significantly decreased, regardless

of the drying method [72] High hydrostatic pressure treatment of orangejuice combined with moderate heat treatment led to an increase inthe extraction of provitamin A carotenoids [73] This was attributed tothe denaturation of the carotenoid-binding protein induced by pressure.Lessin et al [8] reported the quantitative distribution of provitamin Aisomers in fresh and processed fruits and vegetables using a polymeric

C30

produce was performed on the same lot On a quantitative basis, cessing produced 50, 26, 22, 19, and 16% increases in total provitamin Acarotenoids in collards, broccoli, sweet potatoes, spinach, and carrots,respectively, relative to the provitamin content in the fresh samples.Since little or no degradation of carotene reportedly occurs duringthermal processing [70,74], these increases are most likely a result ofincreased extraction efficiency due to disruption of carotoprotein com-plexes, inactivation of carotene oxidizing enzymes, and loss of solublesolids into the liquid canning medium For the majority of the samplesanalyzed, the provitamin all-trans isomer was lower in the processedsamples as compared to the fresh samples on a percent basis This isbecause thermal processing causes the conversion of trans isomers to cisisomers Solubilization of carotenes is a prerequisite for the formation ofcis isomers Severe heat treatment causes solubilization of carotenes bycellular lipids released after thermal breakdown of cell structures [75].Canning of sweet potatoes caused the largest increase in total cisisomers (39%), followed by processing of carrots (33%), tomato juice(20%), collards (19%), tomatoes (18%), spinach (13%), peaches (10%)and orange juice (3%) The predominant cis isomer of provitamins in pro-cessed red, yellow, and orange fruits and vegetables was 13-cis, whereas

in green vegetables the predominant isomer of b-carotene (the only vitamin detected) was the 9-cis isomer The boiling of broccoli for 7 mindid not cause significant change in the isomer distribution Chen andChen [76] reported significant isomerization of a- and b-carotene afterboiling carrots for 30 min The trans to cis isomerization of thermally pro-cessed fruits and vegetables results in some loss of provitamin A activitybecause of the lower potency of the cis isomers

pro-The freezing of fruits and vegetables preserves the carotenoids,but increasing the lag time after thawing progressively decreasesstationary phase for HPLC (Table 3.4) Analysis of fresh and processed

the carotenoid content [72] This is due partly to resumption of ase activity The b-carotene content of green beans and broccoli did notchange significantly during the retail-marketing period or during frozen

3.3.3 Vitamin A Equivalency

In 2001, the U.S Institute of Medicine [78] changed the interconversion

of vitamin A and carotenoids from a system based on retinol equivalents(mg RE) to one based on retinol activity equivalents (mg RAE) The retinolactivity equivalency ratio for b-carotene from food is set at 12:1 and 24:1for other provitamin A carotenoids In other words, 1 mg of retinol isnutritionally equivalent to 12 mg b-carotene or 24 mg of other provitamin

A carotenoids The 12:1 ratio for dietary b-carotene is derived from thefollowing experimental observations: (1) 6 mg of b-carotene from amixed diet is nutritionally equivalent to 1 mg of b-carotene in oil; and(2) 2 mg of b-carotene in oil yields 1 mg of retinol The 24:1 ratio forother provitamin A carotenoids is based on the observation that thevitamin A activity of b-cryptoxanthin and a-carotene is approximatelyhalf of that for b-carotene The RE and RAE systems are compared inTable 3.8 Food composition data tables should report food content inamounts of each carotenoid whenever possible

Based on total liver vitamin A stores in Mongolian gerbils, the retinolequivalency for 9-cis-b-carotene was 38%, and that for 13-cis-b-carotenewas 62% that of all-trans-b-carotene [79]

The new retinol activity equivalency ratio of 12:1 for b-carotene in amixed diet was based on data from healthy people in developedcountries Field studies conducted in Indonesia and Vietnam indicate alower ratio of 21:1 This implies that populations in developing countriesTABLE 3.8

Comparison of Equivalency Factors for Interconversion of Vitamin A andProvitamin A Carotenoids

¼2 mg of supplemental all-trans-b-carotene ¼2 mg of supplemental all-trans-b-carotene

¼6 mg of dietary all-trans-b-carotene ¼12 mg of dietary all-trans-b-carotene

¼12 mg of other dietary provitamin A

are unable to meet their vitamin A requirements from existing dietarysources [80].

3.3.4 Applicability of Analytical Techniques

Since the mid-1970s till the present day, the method of choice for mining vitamin A and the provitamin A carotenoids in food has beenHPLC In the 1995 edition of Official Methods of Analysis of AOAC Inter-national, HPLC methods have been introduced for the first time for the deter-mination of vitamin A in milk [81] and milk-based infant formulas [82]

deter-Of the vitamin A commonly found in foods, only all-trans-retinol andsmaller amounts of 13-cis-retinol, both in esterified form, are usuallypresent in significant quantities For the analysis of vitamin A-fortifiedfoods, HPLC can be applied to determine either the total retinol content

or the added retinyl ester (acetate or palmitate), depending on theextraction technique employed

The vitamin A activity of plant foods is usually based on the HPLCdetermination of the three most ubiquitous provitamins, namely a- and

provita-mins from other carotenoids and to quantify them individually Anobvious prerequisite to accurate quantitation is the conclusive identifi-cation of the provitamins

To assess the effects of processing on the nutritional value of a plantfood with respect to vitamin A activity, the various isomeric forms of pro-vitamin A carotenoids present in both the fresh and processed states must

be accurately measured In such investigations it must be demonstratedthat the analytical procedure does not itself cause trans –cis isomerization

of carotenoids

The HPLC methodology for carotenoids depends on their knowndistribution in plant tissues, which can be classified into three maingroups: (1) those in which the vitamin A value is due almost exclusively

to b-carotene (e.g., green leafy vegetables, peas, broccoli, sweet potatoes,tomatoes, water-melon, mango); (2) those in which primarily a- and

of squash); and (3) those in which b-cryptoxanthin and b-caroteneare the major contributors (e.g., cashew, apple, peach, persimmon,loquat) [83]

3.4 Intestinal Absorption, Metabolism, and Transport

The following discussion of absorption, metabolism, and transport istaken largely from a more detailed account by Ball published in 2004 [84]

3.4.1 Absorption

The absorption of b-carotene and retinol and principal metabolic eventswithin the enterocyte are shown diagrammatically in Figure 3.3 [85].Absorption of vitamin A and carotenoids depends on the proper function-ing of the digestion and absorption of dietary fat, which are described

in Section 2.2.4 Ingested retinyl esters and provitamin A carotenoidsare liberated from their association with membranes and lipoproteins

by the action of pepsin in the stomach and of proteolytic enzymes inthe small intestine In the stomach, the free carotenoids and retinylesters congregate in fatty globules, which then pass into the duodenum.Extensive hydrolysis of retinyl esters takes place within the duodenum,catalyzed mainly by a nonspecific pancreatic hydrolase that can act on

a wide variety of esters Retinyl ester hydrolysis is completed by border hydrolases Esterified xanthophylls are hydrolyzed by esterases

brush-In the presence of bile salts, the fatty globules are broken up intosmaller globules, which renders them more easily digestible by avariety of pancreatic lipases Above a critical concentration of bile salts,

central cleavage retinaldehyde

fatty acids

retinoic acid INTESTINAL LUMEN

ENTEROCYTE bile

2-50%

FIGURE 3.3

Intestinal absorption and metabolism of b-carotene and retinyl esters (Reproduced from Erdman, J.W., Jr., Bierer, T.L., and Gugger, E.T., Ann N.Y Acad Sci., 691, 76, 1993 With permission.)

the bile constituents form micelles The lipolysis products, together withretinol and carotenoids, combine with the micelles to form mixed micelles.The retinol and carotenoids contained within the mixed micelles crossthe unstirred layer of the intestinal lumen and are released as a result ofmicelle dissociation in the brush-border region Physiological concen-trations of retinol derived from natural food sources are absorbed byfacilitated diffusion, mediated by a carrier; at higher concentrations, aprocess of simple diffusion takes over The carrier-mediated absorption

of retinol shows specificity toward all-trans-retinol; uptake of 9-cis- and13-cis-retinol and retinaldehyde takes place by simple diffusion [86]

In vitro studies in Caco-2 cells have suggested that carotenoid uptake byenterocytes is a facilitated process [87]; however, no transporter proteinshave as yet been identified for carotenoids

Human ingestion of an algal preparation consisting of a 50:50 mixture

of all-trans- and 9-cis-b-carotene produced only a modest increase in theserum concentration of the 9-cis isomer with respect to the increase in theall-trans isomer [88,89] In contrast, ingestion of 13-cis-b-carotene fromnatural palm oil resulted in significant elevation in its plasma concen-tration [90] Stahl et al [91] reported the preferential accumulation ofall-trans-b-carotene in chylomicrons compared with the 9-cis isomer,suggesting an efficient isomer-selective mechanism for intestinal uptake

of b-carotene In stable isotope studies, isomerization of 9-cis to

physio-logic doses, and shown to occur before its secretion into the bloodstream[92] Whereas 13-cis-b-carotene has no known specific function, the 9-cisisomer can be converted to both all-trans- and 9-cis-retinoic acid [93,94],which are hormones involved in the regulation of gene expression[95,96] The isomer-selective ability of the intestinal mucosa may beimportant in limiting the potential supply of 9-cis retinoids to tissues

3.4.2 Metabolic Events Within the Enterocyte

3.4.2.1 Esterification of Retinol

Within enterocytes, retinol becomes bound in a 1:1 molar ratio to cellularretinol-binding protein type II (CRBP-II), which is present exclusively andabundantly in these cells The protein-bound retinol is esterified withsaturated long-chain fatty acids, preferentially palmitic acid (C16:0).The esterification uses a different pool of fatty acids, and hence differentenzymes, than are used for the synthesis of triglycerides

3.4.2.2 Conversion of Provitamin Carotenoids to Retinoids

[97] Both symmetric and asymmetric oxidative cleavage of provitamin

carotenoids take place within the enterocyte [98] Symmetric cleavage at

retinalde-hyde; other provitamin carotenoids yield one molecule of retinaldehyde.The enzyme responsible for the central cleavage of b-carotene, b-carotene-

mouse tissues and found to be expressed not only in duodenal villi,but also in liver and in tubular structures of lung and kidney [99].Redmond et al [100] confirmed that the dioxygenase is a cytosolicenzyme and showed that it is also highly expressed in the testis.Leuenberger et al [101] provided evidence that the central cleavage of

rather than a dioxygenase-catalyzed one

In the excentric cleavage of b-carotene described by Glover [102], onemolecule of b-carotene ultimately yields one molecule of retinaldehyde

0

to retinoic acid is postulated to take place by a b-oxidative-type enzymesystem [103] All of the b-apocarotenals formed from b-carotene can

15 15′

be shortened to retinaldehyde Kiefer et al [104] cloned a mammalian

Most of the retinaldehyde formed from carotenoids becomes bound toCRBP-II and reversibly reduced to retinol by retinaldehyde reductase[105] The resulting retinol– CRBP-II complex is then used as a substratefor esterification

Bioconversion of provitamins to retinoids is regulated both up anddown according to vitamin A status The activity of the dioxygenasewas shown to be higher in rats [106] and in hamsters [107] fed avitamin A-deficient diet compared with normally fed controls This up-regulation was confirmed in rats by van Vliet et al [108], who alsofound that a high intake of either retinyl ester or b-carotene down-regulated (decreased) cleavage activity Using a stable isotope-dilutionprocedure to indicate total body stores of vitamin A, Ribaya-Mercado

et al [109] showed that bioconversion of plant carotenoids to vitamin A

in Filipino school-aged children increases when vitamin A status is low

In the dioxygenase assay, the 9000 g supernatant (S-9, cell fraction ofintestinal mucosa with the highest cleavage activity) is incubated withthe carotenoid of interest and the retinaldehyde yield is measured.Using this in vitro measurement of cleavage activity, retinaldehydeformed from a-carotene and b-cryptoxanthin was 29 and 55%, respec-tively, of the amount formed from b-carotene Addition of 9 mg of lutein

to an incubation with 3 mg b-carotene reduced retinaldehyde formation,while lycopene had no effect [110] Many green vegetables (e.g., broccoli,kale, spinach, watercress, Brussels sprouts, green beans, and peas) containmore lutein than b-carotene [30,111,112] and so a predominant vegetablediet might lower the provitamin A activity of b-carotene in vivo Thiscould at least partly account for the vitamin A deficiency found in somedeveloping countries, in spite of adequate intake of b-carotene

Dioxygenase activity is also affected by dietary protein and fat Thereduction in dioxygenase activity in response to protein insufficiency[113] may account for the reduced conversion of b-carotene to vitamin

A in the intestinal wall of rats fed low-protein diets [114] During et al.[115] reported that feeding rats a diet rich in polyunsaturated fattyacids (PUFA) (basal diet with 15% soybean oil) enhanced the activity of

C O

15 15′

intestine This dual enhancement was not observed for a diet rich in rated fatty acids (basal diet with 15% hydrogenated soybean oil) The datasuggested that the carotene-cleaving enzyme and retinol-binding proteinare regulated by a common mechanism involving PUFA CRBP-II isinvolved in the conversion of retinaldehyde to retinol, therefore the sim-ultaneous increases of the enzyme activity and protein level wouldenhance the total ability of intestine to convert b-carotene to vitamin

satu-A In support of this hypothesis, the feeding of Mongolian gerbils with

a diet providing high amounts of PUFA resulted in higher liver vitamin

A and lower liver b-carotene levels [116]

3.4.3 Liver Uptake of Chylomicron Remnants and Storage

of Vitamin A

The combined retinyl esters, accompanied by varying small amounts ofunchanged carotenoids, are secreted from the enterocytes into the blood-stream via the lymphatic system as components of chylomicrons [117].The plasma concentrations of chylomicrons typically peak at 4 –5 h after

a meal [118] The circulating chylomicrons undergo lipolysis and theresultant chylomicron remnants are taken up by the liver and to a lesserextent by extrahepatic tissues Within the liver, the chylomicron remnantretinyl esters are hydrolyzed and most of the retinol is transferred fromhepatocytes to stellate cells, where it is esterified and stored

The amount of vitamin A stored in the liver influences retinol utilization

by extrahepatic tissues, and therefore hepatic liver reserves are a trueindication of vitamin A status Green et al [119] determined the retinol utili-zation rate in rats provided with different intakes of vitamin A, such that therats had low, marginal, or high liver vitamin A reserves Vitamin A-depletedrats exhibited a lower utilization rate, which was positively correlated withthe size of the plasma retinol pool; that is, the lower the plasma retinolconcentration, the lower the vitamin A utilization rate The increased rate

of utilization observed in rats of higher vitamin A status was reflected

in an increased rate of retinol catabolism It appeared that some minimalutilization rate is maintained as long as dietary supply and/or liver stores

of vitamin A can maintain normal plasma retinol concentrations Thedecreased utilization rate in depleted states could be a way of conservingvitamin A for its most critical functions, whereas in vitamin A sufficiencyincreased catabolism prevents excessive accumulation of retinol Acceler-ated catabolism as a function of increase in liver vitamin A stores was alsoreported in rats fed excessive amounts of vitamin A [120]

3.4.4 Plasma Transport of Retinol and Carotenoids

Upon demand, the retinyl esters in the liver are hydrolyzed and theretinol is released into the bloodstream bound to retinol-binding protein

(RBP), which is synthesized in the liver In the plasma, retinol – RBP bines with a protein called transthyretin, and the resultant complex deliversretinol to vitamin A-requiring cells Plasma concentrations of vitamin Aare maintained relatively constant through a homeostatically controlledinterchange of the vitamin among plasma, liver, and extrahepatic tissues.Remnant carotenoids do not accumulate in liver cells: they are releasedinto the circulation as components of very low density lipoproteins(VLDL), which are subsequently delipidated to low density lipoproteins(LDL) In the fasted state, LDLs are the main carriers of nonpolar caroten-oids Carotenoids released from lipoproteins, especially LDL, are taken up

com-by many tissues, particularly adipose tissue, where they accumulate [121].Rock et al [122] examined the plasma carotenoid response to a diet ofnatural foods that was very low (,4 mg/day) in carotenoids Plasmacarotenoid concentrations were measured on study days 2– 3, 14 – 15,

b-carotene occurred between days 2– 3 and 14– 15, but did not occurbetween days 14 –15 and 35 –36 Plasma levels on days 63 –64 averaged23% of mean initial levels The initial rapid plasma response to a low-carotenoid diet suggests that measurement of plasma carotenoid levels

is useful only in the assessment of short-term intake The slower rate ofdecline after 2 weeks suggests that, during dietary carotenoid depri-vation, carotenoids are released from tissues back into plasma Regularconsumption of carotenoid-containing foods seems to be necessary tomaintain plasma levels of these compounds

3.4.5 Tissue Uptake and Metabolism of Retinol

The RBP component of the circulating retinol complex is recognized by areceptor on the surface of target cells and, after negotiating the lipid layer

of the plasma membrane, the retinol interacts with a specific cellularbinding protein Cytoplasmic retinol is subject to a variety of metabolicfates, including oxidation to retinoic acid Cellular binding proteinsplay major roles in controlling retinol metabolism and may regulate themovement of retinoic acid to the nucleus, where it acts as a hormone toinfluence gene expression

3.5 Bioavailability

3.5.1 Introduction

The published literature on carotenoid bioavailability is extensive, butcomprehensive reviews are available [98,123,124] To simplify matters,

this discussion focuses on the major provitamin A carotenoid, b-carotene.Bioavailability is an ambiguous term and therefore it is important tostate the method when assigning a bioavailability value to a food orsupplement Bioavailability of carotenoids is frequently assessed byabsorption efficiency, which is defined as the percentage of ingested caro-tenoid that is secreted into the general circulation and therefore madeavailable for tissue uptake [124] Two related terms have been introducedinto the carotenoid literature, bioaccessibility and bioefficacy Bioaccessi-bility is defined as the fraction of carotenoid transferred during digestionfrom the food matrix to mixed micelles and thus made accessible forabsorption [125] Bioefficacy defines the percentage efficiency withwhich ingested b-carotene is absorbed and converted to retinol inthe body [126] Because 1 mmol b-carotene theoretically could form

2 mmol retinol, 100% bioefficacy would mean that 1 mmol dietary

yielding 2 mmol retinol (0.572 mg) Thus, the amount of b-carotenerequired to form 1 mg retinol would be 0.537/0.572 ¼ 0.94 mg

When meals containing natural amounts of vitamin A and provitamin

A carotenoids are consumed, vitamin A is absorbed with an efficiency of

70 –90% compared with 20 –50% for the provitamins The absorptionefficiency of vitamin A remains high (60 – 80%) as the amount ingestedincreases beyond physiological levels, whereas that of the provitaminsfalls dramatically to less than 10% [127], indicative of a saturableprocess When a large dose (40 mg) of b-carotene (in the form of capsulescontaining palm oil carotenes) was administered with a standardized,low-carotenoid meal to men and women, only 3.5% of the dose appeared

in the chylomicron-rich fraction of plasma [128] Tyssandier et al [129]reported a very low (,7%) recovery of vegetable-borne carotenoids inthe micellar phase of the duodenum (bioaccessibility), which couldaccount for their poor bioavailability

3.5.2 In vivo Methods of Assessing b-Carotene Bioavailability

3.5.2.1 Use of Radioisotopes in Cannulated Patients

In two studies designed to investigate the intestinal handling ofb-carotene in humans [130,131], single doses within the range of 45 mg

liquid formula were administered to four male hospitalized patientswith cannulated thoracic ducts Radioactivity was measured in four frac-tions of the lymph; namely, b-carotene, retinyl ester, retinaldehyde, andretinol fractions Recoveries of radioactivity in the lymph of three patientswere 9, 15, and 17% Of the recovered radioactivity in these patients, 68, 73,and 88% was found in the retinyl ester fraction, and 1.7, 11.3, and 27.9% was

found in the b-carotene fraction In the fourth patient, 52% of the tered radioactivity was recovered in the lymph, the majority (90%) ofthis occurring in the b-carotene fraction These studies clearly showedthat, apart from the fourth patient, humans are moderate absorbers

adminis-of intact b-carotene, and bioconvert b-carotene primarily to retinyl esters

3.5.2.2 Animal Models

The use of radioisotopes in healthy human subjects cannot be justifiedbecause of the risks involved Animal models offer the advantages ofallowing use of radiolabeled compounds and accessing tissues Ratsabsorb b-carotene in a different manner to humans and therefore, arenot appropriate models for studying carotenoid absorption and bioavail-ability When rats are fed radioactive b-carotene, the b-carotene is con-verted almost entirely to retinyl esters at the level of the enterocyte, andthere is no labeled b-carotene in the lymph [132] This means that,unlike in humans, accumulation of b-carotene in rat tissues does notoccur and neither does it occur in the tissues of mice, hamsters, guineapigs, rabbits, chickens, pigs, or sheep In the search for a more suitableanimal model, the ferret and gerbil have been found to resemble thehuman in their ability to absorb and accumulate intact carotenoids[133] Gerbils are the more suitable because they are small (an adultmale weighs 80 g) and easily maintained in large numbers In addition,gerbils convert dietary b-carotene to vitamin A with efficiency similar tothat currently estimated for humans Ferrets are large enough (an adultmale weighs 1200 g) to permit surgical cannulation of the lymphaticsand the portal vein, allowing direct measurement of carotenoid absorp-tion Preruminant calves also resemble humans in their ability to absorb

blood and liver biopsy samples

3.5.2.3 Serum, Plasma, or Chylomicron Responses not Involving

Isotopic Tracers

Plasma retinol levels are homeostatically maintained within a narrowrange (except in cases of vitamin A deficiency or excess) and so do notreflect dietary intake of either vitamin A or provitamin A carotenoids inadequately nourished subjects This explains why serum vitamin Alevels were unchanged when human subjects were fed carrot juice[134] Plasma or serum carotenoid responses (concentration versus timecurves) do reflect dietary intake of carotenoids The relative bioavailabil-ity of b-carotene in a particular food or meal has been determined as theserum b-carotene response relative to that after simultaneous ingestion ofpure b-carotene and a low-carotenoid meal

Average serum b-carotene concentrations in women increased 4.1- and4.0-fold in response to single oral doses of 60 and 210 mg b-carotene,respectively Maximum serum b-carotene levels were reached 24 h post-dosing and, after 8 days, serum levels remained ca twofold higher thanbaseline levels [135] Increases in serum b-carotene concentrationsvaried from 3.5- to 11-fold (mean: 7.4-fold) among women after a28-day period of supplementation with 30 mg purified b-carotene [136].These increases conformed to first-order kinetics with a half-time of5.5 days.

An increase in dietary intake of fruits and vegetables is associatedwith an increase in plasma carotenoid concentrations [137 – 141] Theplasma response is very variable in magnitude and duration, and somepeople consistently exhibit a low response [142] Johnson and Russell[143] measured b-carotene concentrations in plasma and various lipo-proteins in healthy males for ten days after a single oral dose (120 mg) ofb-carotene in capsule form Seven of the eleven subjects were nonrespon-ders, showing little or no increase in plasma b-carotene and only a smallresponse in chylomicrons Of interest with the four responders was thatsurges of chylomicron b-carotene occurred every few days followingthe single dose, suggesting delayed release of b-carotene from enterocytes

or re-uptake of b-carotene following sloughing-off of enterocytes.Density gradient ultracentrifugation isolates a triglyceride-rich lipo-

mainly intestinally derived chylomicrons and their remnants togetherwith some VLDL of hepatic origin Measurement of postprandialchanges in the concentration of retinyl esters in the TRL fraction providesinformation as to the mass of vitamin A derived from a test meal contain-ing provitamin A carotenoids The value of this approach stems from thefact that chylomicrons and their remnants are the exclusive means bywhich newly absorbed b-carotene and derived retinyl esters are trans-ported from the intestine to the liver In the postprandial chylomicron(PPC) response model, using human subjects, data from multiple bloodsamples drawn over periods ranging from 8 to 16 h after a test meal areused to construct concentration versus time plots, from which area-under-the-curve (AUC) values are computed Such values are presumed

to be proportional to the amount of retinyl ester produced by the nal mucosa [144]

intesti-Response curves in whole human plasma or serum do not measureaccurately the extent of absorption of carotenoids from a test mealbecause they are unable to distinguish newly absorbed carotenoidsfrom endogenous carotenoids circulating within LDL Response curves

in TRL fractions specifically measure newly absorbed carotenoids, aswell as accounting for intestinal conversion of provitamin A carotenoids

to retinyl esters Another advantage of TRL response curves is the

allowance for smaller doses that results from lower baseline levels.van Vliet et al [145] reported the following observations supporting thepresence of mainly intestinally derived lipoproteins in the TRL fraction:(1) the TRL fraction from fasting plasma contained no b-carotene; (2)control subjects showed no response or only a very low response; and(3) the increase of plasma b-carotene seen up to 16 h after the first meal(probably due to liver-derived lipoproteins) was not seen in the TRL frac-tion Hu et al [146] subfractionated TRLs into particles of varying size andcomposition in an effort to better distinguish particles of exogenous andendogenous origin.

3.5.2.4 Methods Involving Stable Isotopes

Parker and colleagues [147] adapted the PPC model by administering

simul-taneously with the test meal The labeled compound constituted anextrinsic reference with which to measure the mass of unlabeled retinyl

carotenoids and the mass of intact carotenoid (as b-carotene) absorbed.The extrinsic reference also controlled for variation in in vivo chylomicronkinetics and for variation in retinyl ester recovery during the preparationand analysis of the TRL fraction The retinyl ester-derived retinol

from HPLC and derivatized to a trimethylsilyl (TMS) ether The molar

was multiplied by the concentration of total TRL retinyl ester, as

the TRL fraction at each sampling time point Baseline-corrected response

subject given a test meal of raw carrot with 20 g fat An absorption

pro-vitamin A carotenoids in the test meal was calculated by dividing the

quotient by 4.8 The same approach was used to estimate the mass of

spinach in amounts that contained 6 mg b-carotene, and 20 g of addedfat The amount of vitamin A reaching the bloodstream after consumption

Thus, in the presence of ample fat, the amount of b-carotene in rawcarrot or raw spinach required to form 1 mg retinol was 20 mg (0.3 mg